Sounding checks proposed decisions and returns structured proceed, revise, or pause assessments.
Scored across 1 tool
There is only a single tool, so there is no possibility of overlap or misselection. Its purpose (bounded decision assessment returning proceed/revise/pause) is clearly stated and distinct.
The single name sounding_decision_check is clear snake_case and readable, prefixed with the server namespace. With only one tool there is no cross-tool pattern to judge, though it follows a noun_verb_check shape rather than a strict verb_noun pattern.
A one-tool server is on the thin side for the rubric's preferred 3-15 range and offers no companion operations. It is a deliberately focused, single-purpose assessment endpoint rather than a sprawling surface, so it is borderline rather than wrong.
For a 'sounding a decision' domain, one well-specified assessment tool plausibly covers the core need, and it explicitly takes rich optional context (evidence, alternatives, constraints, reversibility). Minor gaps exist—no way to retrieve or compare prior decisions across calls and no batch mode—but agents can work around these by resubmitting context.
A bounded decision assessment: proceed, revise, or pause. Supply decision (or proposed_action) and goal. More relevant context supports better guidance: include the evidence for and against, alternatives considered, earlier decisions or rules that bear on the choice, constraints, affected people, reversibility, urgency and decision owner. Sounding sees only what you send. Optional means not required to submit, not unnecessary to assess. Preserve unknowns; never invent facts or authority. Takes no external action.
| Name | Required | Description | Default |
|---|---|---|---|
| goal | Yes | The outcome you want to achieve. | |
| task_id | No | Optional reference; generated when omitted. | |
| urgency | No | Actual timing needs and their reason. | |
| decision | No | The proposed action, options or question to assess. Use decision or the compatible proposed_action field. | |
| evidence | No | Facts that support the action, one per item. Put facts against it in evidence_against. | |
| constraints | No | Limits, requirements and permission boundaries. | |
| alternatives | No | Other options considered, including doing nothing or a smaller step, each with one line on why it was set aside. | |
| reversibility | No | How easily the action can be undone, and what cannot be undone. | |
| decision_owner | No | Who makes the final decision. | |
| affected_people | No | Who is affected, including people not in the conversation. | |
| prior_decisions | No | Earlier decisions, rules, definitions or standards in your project that bear on this choice, stated as they stand. Sounding cannot see your project. | |
| proposed_action | No | Compatible alternative to decision. If both are supplied, they must match. | |
| evidence_against | No | Facts, risks and costs that count against the action, including side effects such as extra work, length or delay. Include them even if you still favor the action. |
| Name | Required | Description |
|---|---|---|
| reason | Yes | |
| task_id | Yes | |
| next_step | Yes | |
| disposition | Yes | |
| evidence_basis | Yes | |
| material_question | Yes | |
| assessment_support | Yes | |
| assumptions_to_check | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Beyond the readOnlyHint annotation, the description adds genuine behavioral context: 'Takes no external action' confirms the consultative nature, and 'Sounding sees only what you send' plus 'never invent facts or authority' disclose the tool's epistemic limits and that nothing is fetched from the surrounding project. It does not cover latency, limits, or output shape, but the output schema carries the latter.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The outcome verdict is front-loaded, and the long list of context fields is necessary given 13 parameters. It is dense and slightly repetitive ('More relevant context supports better guidance' followed by an enumerative expansion of that same idea), but every sentence carries information and nothing is clearly disposable.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
With annotations covering safety, a 100%-covered schema, and an output schema handling return values, the description's remaining job is to convey the consultative posture and input expectations, which it does thoroughly. The one real gap is the absence of any when-to-use framing, which is minor for a solo-source tool with no siblings.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 100%, so the baseline is 3; the description earns an extra point by explaining the semantics of optionality ('Optional means not required to submit, not unnecessary to assess') and by framing why the optional context fields matter for output quality. This is meaning beyond the schema, not a restatement of it.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The opening sentence names the operation and its three possible outcomes (proceed, revise, pause), so an agent knows this is a decision-assessment tool that returns a bounded verdict. It is clear and specific, though 'bounded decision assessment' is abstract enough that the value is mostly conveyed by the outcome list rather than a plain verb+resource statement.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description spends its guidance budget on what to supply ('include the evidence for and against, alternatives considered...') rather than when to call the tool versus not calling it. There are no sibling tools to route against, but neither is there any stated trigger condition or exclusion. Usage is implied by the input recipe rather than stated.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
sounding_decision_check9 fields changedInput schema / properties / affected_people / descriptionAdded value: +"Who is affected, including people not in the conversation."Input schema / properties / alternativesAdded value: +{
+ "description": "Other options considered, including doing nothing or a smaller step, each with one line on why it was set aside.",
+ "items": {
+ "type": "string"
+ },
+ "type": "array"
+}Input schema / properties / constraints / descriptionAdded value: +"Limits, requirements and permission boundaries."Input schema / properties / decision_owner / descriptionAdded value: +"Who makes the final decision."Input schema / properties / evidence / descriptionAdded value: +"Facts that support the action, one per item. Put facts against it in evidence_against."Input schema / properties / evidence_againstAdded value: +{
+ "description": "Facts, risks and costs that count against the action, including side effects such as extra work, length or delay. Include them even if you still favor the action.",
+ "items": {
+ "type": "string"
+ },
+ "type": "array"
+}Input schema / properties / prior_decisionsAdded value: +{
+ "description": "Earlier decisions, rules, definitions or standards in your project that bear on this choice, stated as they stand. Sounding cannot see your project.",
+ "items": {
+ "type": "string"
+ },
+ "type": "array"
+}Input schema / properties / reversibility / descriptionAdded value: +"How easily the action can be undone, and what cannot be undone."Input schema / properties / urgency / descriptionAdded value: +"Actual timing needs and their reason."sounding_decision_check3 fields changedOutput schema / properties / assessment_supportAdded value: +{
+ "additionalProperties": false,
+ "properties": {
+ "explanation": {
+ "pattern": "\\S",
+ "type": "string"
+ },
+ "level": {
+ "enum": [
+ "strong",
+ "moderate",
+ "tentative"
+ ],
+ "type": "string"
+ }
+ },
+ "required": [
+ "level",
+ "explanation"
+ ],
+ "type": "object"
+}Output schema / properties / confidenceRemoved value: -{
- "enum": [
- "high",
- "low",
- "medium"
- ],
- "type": "string"
-}Output schema / requiredPrevious value: -[
- "task_id",
- "disposition",
- "reason",
- "material_question",
- "evidence_basis",
- "assumptions_to_check",
- "next_step",
- "confidence"
-]New value: +[
+ "task_id",
+ "disposition",
+ "reason",
+ "material_question",
+ "evidence_basis",
+ "assumptions_to_check",
+ "next_step",
+ "assessment_support"
+]sounding_decision_check9 fields changedInput schema / anyOfAdded value: +[
+ {
+ "required": [
+ "decision"
+ ]
+ },
+ {
+ "required": [
+ "proposed_action"
+ ]
+ }
+]Input schema / properties / decisionAdded value: +{
+ "description": "The proposed action, options or question to assess. Use decision or the compatible proposed_action field.",
+ "pattern": "\\S",
+ "type": "string"
+}Input schema / properties / goal / descriptionAdded value: +"The outcome you want to achieve."Input schema / properties / goal / patternAdded value: +"\\S"Input schema / properties / proposed_action / descriptionAdded value: +"Compatible alternative to decision. If both are supplied, they must match."Input schema / properties / proposed_action / patternAdded value: +"\\S"Input schema / properties / task_id / descriptionAdded value: +"Optional reference; generated when omitted."Input schema / properties / task_id / patternAdded value: +"\\S"Input schema / requiredPrevious value: -[
- "task_id",
- "goal",
- "proposed_action",
- "evidence",
- "constraints",
- "affected_people",
- "reversibility",
- "urgency",
- "decision_owner"
-]New value: +[
+ "goal"
+]sounding_decision_checkCheck AI work against requirements and return structured verdicts, findings, and repair steps.
Evidence-weighted decision engines: AI business cases, venture readiness, free soundchecks.
Log, evaluate, and ground AI decisions against authority context. Returns PASS, WARN, or BLOCK.
Deterministic decision layer for autonomous agents: reproducible PROCEED/REVIEW/SKIP verdicts.
Add one secure layer between your agents and this server.
